the first theme isolated in its every version. Our conclusions accord better with his further statements. He asserts that the Fabulae Extravagantes text=(S) of Cock and Fox agrees completely in essentials with Ba and BR; which supports our derivation of these from a common source. He thinks that Ad cannot go back to antiquity; and our hypothesis sends it back to antiquity only for its first part. He considers M the best and most natural version. Finally he holds that the version known to M was that which served the need of the composers of Renart and the other epics, including Chaucer. This is going a remove farther back than we had gone: the one supposition seems quite as tenable as the other.

2. Voretsch, "Der Reinhart Fuchs und der Roman de Renart," in Zeitschrift für romanische Philologie, Vol. XV, pp. 136-47. He considers the Wolf stories as constituting one division of the fable. He believes that Chaucer comes out of Renart, Branch II, directly, but claims that it is widely different from the original. He observes that in the Renart as in GS the cock closes first one eye, then both. This cannot be the invention of a trouvère; therefore it is probably an addition to a reworking of Reinike Fuchs. For us, this shows still more clearly the relation between fable and epic.

3. Du Méril, Poésies inédites du moyen âge, pp. 215, 216, has some conjectures concerning Baldo, whose versification he considers too elaborate for the eleventh century, and not sufficiently developed for the thirteenth-the latter point being also supported by external evidence.

It would be a very precious fact for literary history, if one could succeed in establishing it by proofs of a more precise date: for most of these fables are imitated from Calilah and Dimnah, and it would result therefrom that the influence of the Orient upon the literary ideas of the Romance peoples had made itself felt earlier than is supposed. The last reflection does not concern the Cock and Fox.

4. G. Paris, "Les fabulistes latins, par Hervieux," in Journal des Savants, 1884, pp. 684, 685, supports Warnke in assigning Ademar's fable to a mediæval source. "The question [of Ad's origin in Phædrus] is much more doubtful for Perdix et

Vulpes,' where the ideas and the style of the Middle Ages seem to rule."

5. Skeat, Complete Works of Geoffrey Chaucer. "The House of Fame, etc. . . . . Account of the Sources of the Canterbury Tales," second ed. (Oxford, 1900), pp. 431, 432, says of the Nonne Prestes Tale:

An early version of the tale occurs in a short fable by Marie de France, afterwards amplified in the old French Roman de Renart. The corresponding portion of the Roman de Renart contains the account of the Cock's dream about a strange beast, and other particulars of which Chaucer makes some use.

According to him, again, M> Renart > Ch.

6. Miss Petersen, Sources of the Nonne Prestes Tale, passim. We fall back on this excellent monograph, as giving perhaps the most elaborate discussion of the Chaucer question, and as raising incidentally, several points bearing on the fable. Miss Petersen contributes these suggestions:

a) Chaucer is "unmistakably epic," as evinced by the features of the dream, the proper names, the description of the cock's owner and of the yard, the dialogue between cock and hen, the lament of the hens-all peculiar to the epic versions. Chaucer's immediate source is "some epic tale belonging to the Renart cycle." (P. 9.)

b) She cites the opinion of Sudre that "the intervention of the dogs. . . . is a survival of the original cadre of the story. This cadre, he thinks, is to be found in the Æsopic fable of the Dog and Cock." She admits that "in the Esopic account, the part of the dog is of great consequence his role as protector

is really the turning-point of the story." But she holds that in the Chanticleer episode the pursuit by the dogs is merely an "accessory theme," and adds with apparent justice, that it may have been "formulated from the observation of real life." Yet she grants the similarity of our Esopic Wolf and Kid story as to the Pursuit theme. (Pp. 10-16.)

1 It may be noted that in the figures around the Bayeux tapestry-which some suppose to derive from Ad-our fable occurs more frequently than any other. See Bruce, The Bayeux Tapestry Elucidated (London 1856), Plates I, II, VI, XIII. It occurs as the Cheese story twice, as Fox and Partridge once, as Cock and Fox perhaps once or twice. This serves well to illustrate the great popularity of the fable.

c) She mentions also the story of the Cat with One Trick, in which the dogs appear, and believes that from some such mediæval "floating tale" the theme of the pursuit by the dogs was drawn and appended to the Cock and Fox Story." (Pp. 18-21.)

d) She follows Warnke in considering Al the originative form; but she wisely differentiates the oculis clausis trick from the first theme, and is right in declaring that this trick itself is not found alone. (P. 46.)

It would take us too far afield to discuss all of Miss Petersen's views. Suffice it to say that she does not actually confute our E hypothesis, and that her Chaucer descent agrees with our table— except that she leans to the belief that the folklore story of Cock and Fox, rather than any special fabular version of it, as M, contributed to the Renart cycle. (For her conclusions see pp. 46, 118.)

7. Furnivall, Origin and Analogues of the Canterbury Tales, p. 115, claims an English origin for the fable. This view is unsupported by others, but seems to me quite tenable, when we remember that the bulk of our versions are more or less directly English-that all save four or five derive directly from Alfred.

We see thus that the Cock and Fox fable has been variously oriented as Esopic or Phædric, popular, clerical, English. Our composite hypothesis admits all of these influences. That is to say, we refer the fable for one part to Phædrus, and for the other to the folk-tale (?) of the wolf. It is possible that in this latter we are to see an English clerical presentation, transmitting its marks to E, which gave on the one hand the partridge story, on the other GS and X. This X remains the secondary source out of which proceed all later versions. The story loses then its clerical character, but maintains its English dominance, becomes finally a regular fable, deviates into the epic, but persists in the end as a crystallized exemplum with a definite history, having evolved out of a mass of chaotic and apparently uncoördinated tales. E. P. DARGAN

JOHNS HOPKINS UNIVERSITY

A VALUABLE MIDDLE ENGLISH MANUSCRIPT

In my search for Old and Middle English versions of the Gospel of Nicodemus I recently came upon a very interesting MS in the Worcester Cathedral Library. And it is a MS which has not thus far attracted the attention of students of English literature and history.' There are at least two reasons why the volume has remained unknown: (1) there is no complete and reliable catalogue of the Worcester collection; (2) the MS, being comparatively late (last quarter of the fifteenth century), and of unattractive appearance generally, would hardly appeal to the average "skimmer" of libraries and seeker after antique treasures. The MS is full of important historical and literary documents, but it is nevertheless entirely ignored in the Historical Commission's report on the Worcester libraries. Nor have I been able to find anything about MS fol. 172 in any of the archæological histories of the city of Worcester.

The MS originally contained at least 226 paper leaves (probably more), of which 16 have been lost from the beginning. So there remain 210 leaves and 6 fly-leaves, 3 at the beginning and 3 at the end, and f. 4 of the modern pagination agrees with f. XVII of the earlier. The MS is bound together in quires of 12 leaves each-except the first quire which has only 6 (and the 3 fly-leaves) the ends of the quires always being indicated by catch-words. The leaves measure 11x8 inches and more than half of them have been considerably injured—perhaps by moisture or heat. So it is difficult to make out the reading of the upper part of the first few pages. The MS is generally without ornamentation, except the original rubrics and capitals in red. The Psalter, however, contains red and blue script in great profusion. One scribe seems to have been responsible for the copying

1 Professor A. S. Napier kindly called my attention to this version of the Gospel of Nicodemus some years ago.

2 Cf. Report of Historical Commission for the year 1895. H. Schenkl has not yet published an account of the Worcester Cathedral Library in bis series of articles on the patristic literature in English libraries. Cf. Wiener Sitzungsberichte since about the year 1890. 67]